The robot that has been working for twenty years

A surgical robot does not operate. It is a set of arms that reproduce the motion of a surgeon's hands at smaller scale, with tremor filtered out and wrists that can rotate further than a human wrist can. The surgeon sits at a console a few metres away and does every part of the procedure. The word robot, here, means a very good remote instrument.

That modesty is the reason it worked. The system never had to be trusted with a decision, so it never had to clear the regulatory and cultural bar that autonomy would require.

What it actually buys

The clinical case rests on minimally invasive access. Operating through several small ports instead of one large incision means less blood loss, shorter stays and faster recovery for the patient, and a dramatically better ergonomic position for the surgeon, who sits rather than stands hunched for four hours. The second of those is underrated and is a genuine reason surgeons ask for the systems.

The evidence on outcomes is more mixed than the marketing. For some procedures, robot-assisted surgery shows clear advantages over open surgery. Against skilled laparoscopic surgery, the advantage is often small and the operating time longer, especially early in a surgeon's learning curve. The honest summary is that it makes difficult minimally invasive procedures accessible to more surgeons, rather than making the best surgeons better.

The business model everyone else wants

A system costs one to two million. That is not where the money is. Each procedure consumes instruments that expire after a set number of uses, and each system carries an annual service contract. Revenue therefore scales with procedures, not with placements, which turns an installed base into an annuity.

This is the model every robotics company would like: sell the machine near cost, earn on consumption. It works in surgery because the consumable is genuinely consumed and regulated. It is much harder in a factory, where a gripper does not wear out on a schedule and a customer can buy a third-party one.

What happened when the patents expired

The core patents lapsed and competitors arrived: large medical device manufacturers, specialist startups, systems aimed at orthopaedics, bronchoscopy and single-port procedures. Prices did not collapse, because the barrier turned out to be the training pipeline, the instrument ecosystem and the hospital's sunk investment rather than the mechanism.

That is a lesson for every robot category. The moat is rarely the machine. It is the ecosystem around it: the people trained on it, the parts supply, the software integrations and the switching cost of retraining a team.

The autonomy question

Automating a surgical step is technically feasible for narrow tasks such as suturing along a marked path, and demonstrations exist. Nothing is approved for autonomous operation on a patient, and nothing will be soon. The regulatory path for a device that makes an irreversible decision on a human body is long, and the liability question has no established answer.

What is arriving instead is assistance short of autonomy: image overlays that show a surgeon where a vessel is, motion limits that prevent an instrument entering a marked region, and analysis of recorded procedures for training. Those improve outcomes without asking anyone to trust a machine with a decision.

What to watch

Watch procedure volume per installed system, which tells you whether new placements are being used or parked. Watch which specialties adopt next, because each one is a separate instrument ecosystem. And watch for the first regulatory approval of any autonomous surgical step, which would be a genuine milestone rather than a demo.